As AI models grow larger and more powerful, they promise incredible capabilities — but at what cost?

Karen Hao is an AI journalist and her new book, Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI, is a New York Times bestseller.

We discuss whether the largest AI models are worth their hefty footprint: They consume massive amounts of electricity and water and Karen argues that smaller models better balance cost vs. benefit.

Karen, who has reported for The Atlantic, MIT Technology Review, and the Wall Street Journal, will also provide a view of AI from outside — far outside — Silicon Valley. She’s reported on AI from across the Global South and says many there feel that AI is a new form of colonialism.

We’ll hear about the fight over data centers in Chile, how New Zealand’s Maori people are using AI to preserve their indigenous language, and why it’s a problem that AI can speak any language, but can only really be policed in a few.

(Our interview was first broadcast in October, while Karen was still writing the book, so we do not discuss her deeply sourced reporting from inside OpenAI.)

—
CRAFTED. is produced by Modern Product Minds, where CRAFTED. host Dan Blumberg and team can help you take a new product from zero to one... and beyond. We specialize in early stage product discovery, growth, and experimentation. Learn more at modernproductminds.com

Subscribe to CRAFTED., follow the show, and sign up for the newsletter 👉 crafted.fm

AI-generated voices aren’t just realistic — they’re changing how brands, creators, and agencies bring ideas to life. In this episode, Wondercraft co-founder Oskar Serrander demonstrates how their “Canva for audio” is unlocking rapid prototyping for high-quality audio ads, content, and storytelling.

You’ll learn:

• How Wondercraft enables you to go from concept to creative in seconds
• Why when AI makes execution easy, ideas and taste matter most
• How brands can test creative faster (and smarter)
• Why audio is still such an under-leveraged medium
• The surprising future of synthetic voice and what it means for storytelling

Oskar also shares his take on where generative AI is heading, why sameness is the enemy of brand, and what this all means for the next generation of creators.

—

Subscribe to the CRAFTED. newsletter 👉 crafted.fm

See how Dan can help you discover, prototype, and launch better products 👉 modernproductminds.com

Kevin Smith is building a totally new kind of podcast app.

Snipd is an AI-native podcast app and building it required a few mindset shifts.

First, what even is a podcast? The way Kevin sees it, podcasts are knowledge. So where most podcast players are, as Kevin calls them, "repurposed music players", Snipd is designed to help you learn. As people listen to episodes, they, or an AI, can save “snips” or interesting moments that they want to remember or share. And the app will also help you review what you've heard, so it reinforces what you've learned.

A second mindset shift is how Kevin had to retrain his engineering brain to build with generative AI. He no longer thinks in if-then-else statements. Rather, he asks himself: How would an intern do it? And not just one intern, but infinite interns…

I learned a ton from the way Kevin thinks and builds, and you will too.

Plus, we discuss the future of podcasting, which looks pretty… weird. You'll talk back to your podcasts, hosts may be synthetic, and shows may not even be designed (at least initially) for human ears.

Chapters:

• (01:30) - Introducing Snipd
• (03:50) - What led Kevin to found Snipd
• (06:10) - How AI changes what's possible with podcasts
• (08:45) - Building with Gen AI requires a mindset shift
• (11:40) - How would an intern solve this?
• (12:35) - How podcast listening and podcasting will change with AI
• (17:35) - Why apps will become your "best friends"
• (22:00) - Why you may talk back to your podcasts

CRAFTED. listeners can try Snipd, and get a free month of the premium version, here.

—
Subscribe to the CRAFTED. newsletter 👉 crafted.fm

See how Dan and Modern Product Minds can help you build better products and level up your product teams.